01 / Introduction

One story, then the four grids

Welcome to the third edition of the Paid Traffic Truth. Every week I take one story from the world of paid traffic and break it down in the Analysis section, then hand you all four grids underneath it. New customer acquisition, first click vs last click, overall channel performance, and new customer lifetime value.

Every number comes from Wicked Reports first party new customer attribution and LTV, aggregated across hundreds of ecommerce brands and verified against real orders. These are the new customer numbers your ad platforms do not show you.

This week's featured story is the one that quietly reshuffles budgets every month. ROAS versus the true cost of a new customer.

02 / Analysis 

05 / Overall Channel Performance

Where the money goes, and THE TRUE COST OF A NEW CUSTOMER 

Meta and Google are about 94% of tracked spend, so this is where the real decisions live. Put the ROAS column next to the nCAC column and read them together. Google's ROAS is roughly three times Meta's, and a new customer costs 39% more on Google. Microsoft posts the best ROAS in the set and the biggest gap between its blended and new customer cost, a 64% markup you never see if you only watch ROAS.

There is a second thing hiding in the aCAC column. On every paid channel the true new customer cost sits above the blended number, because blended quietly includes your existing customers coming back. It is widest exactly where ROAS looks best. That is not a coincidence. The channels that look most efficient are the ones leaning hardest on demand someone else created.

 06 / New Customer Lifetime Value

What a new customer becomes over a year

This is the grid that finishes the story. Meta is the cheapest new customer to acquire and the lowest one year value at $97, a low order value high frequency profile. Microsoft is the most expensive to acquire and the highest one year value at $341. TikTok is expensive on day one but more than doubles its value by the one year mark, the strongest growth curve in the set.

Now the ROAS number makes sense, and it also makes clear why you should not trust it. Microsoft's strong ROAS is really a story about high value customers who close fast. Meta's weak ROAS is really a story about cheap acquisition of lower value customers who need time. Those are two completely different decisions, and ROAS gives you one blurry number for both. One caveat. This blends hundreds of brands at different price points, so read it as a directional market benchmark, not a promise for your store.

07 / Conclusion 

Split the number, then decide

ROAS is not a business metric. It is an efficiency metric for a single platform, and it hides the two things you actually need to run acquisition, the cost of a new customer and the value of that customer over time. The channel that looks best on ROAS was the most expensive place to buy a customer this week. If you had moved budget on ROAS alone, you would have paid more to grow slower.

The fix is not a better dashboard. It is two verified numbers next to each other. What did a new customer cost, and what will that new customer become. Get those and the budget decision stops being a guess.

Every platform grades its own homework. Facebook's ad manager included.

Two things happened this week that every DTC brand running paid media needs to understand. Both are quiet. Neither comes with an alert in your dashboard. And together they explain why your attribution numbers are getting less trustworthy at the exact moment your platforms are making them look more complete.

The Google Signals Fallback Is Gone in a week. Here's What Breaks If You're not ready

What Actually Changes on June 15 

On June 15, Google is changing how Consent Mode works. After that date, ad_storage becomes the sole parameter that governs Google Ads data collection. The workaround most advertisers have been quietly relying on — Google Signals filling in the gaps when website consent is missing — gets discontinued.

What Changed, What It Means, and What to Watch in June

There’s been a lot of chatter this week about Meta offering a paid “elite” subscription tier. Let me cut through the noise and tell you what’s actually happening — and more importantly, what it means for every dollar you’re spending on paid ads right now.

ChatGPT just became an ad platform.

High-intent users. Contextual placement. Buyers in research mode asking questions your ads can actually answer.

If you're a performance marketer, you're already thinking about it.

Here's the problem nobody's talking about : the way you currently track ads wasn't built for this. Plug ChatGPT ad spend into your existing GA4 and Meta setup and you'll get confident-looking numbers that are almost certainly wrong.

Same flawed attribution. New channel. Faster budget burn.

I cover what actually breaks when you try to track ChatGPT ads, and how to set it up so the data underneath is finally honest.

Meta's platform reports a healthy cost per acquisition. Meanwhile, your new customer count is flat. That's not a coincidence — it's a measurement problem. Here's the attribution gap hiding inside your ad account, how to find it, and what to measure instead.  

If your team spends Mondays arguing over dashboards and guessing what to optimize, it’s time for a change. The Five Forces Playbook turns data into decisions with a simple, repeatable weekly routine.